What makes a business worth buying

What makes a business worth buying is a question best answered before you ever look at a single company. I answer it with a written buy-box: a fixed set of business acquisition criteria on one page, so the decision is made on paper long before a founder is across the table and the story starts to feel good.

A gut feel bends. It bends toward the seller you like, the sector that is in the headlines, the deal you have already spent three weeks on and do not want to walk away from. A written filter does not bend. It says yes or no to the same standard every time, and it lets me pass on ninety-nine businesses without regret because the hundredth is the one that fits. Below is what sits on that page, and why each line earns its place.

The buy-box, on one page

Everything below fits on a single sheet on purpose. When the criteria live in your head, they drift; every founder you like reshapes them a little. When they live on paper, they hold. I read the same lines against every business, in the same order, before I let myself get interested in any of them. If a company fails a line, the conversation is over, and I have lost nothing but an afternoon.

Here is the filter as I actually run it. A reader can run it too, against any business in front of them right now.

• Earnings held through the last soft patch, not just the last hot year
• The books reconcile to the bank without a translator
• Ten or more years of trading history behind it
• Demand with a floor: recurring work, a part that wears out, a service nobody can skip
• A clear reason the business still exists in twenty years
• No single customer worth more than a fifth of revenue
• Owner dependence that is fixable, priced into what I pay
• A boring, permanent need rather than a story that has to keep being retold

Profitability that survives a cycle

The first line of the filter is real earnings, not a good year. A business worth buying makes money when the economy is flat and when a competitor undercuts it on price, not only when demand is running hot. I want to see profit that held through the last soft patch, because that is the profit I am actually buying. Anyone can post a strong number in a strong market. The question is what the business does in the year nobody wants to talk about.

Real also means the books can be diligenced. Reported profit is only a useful number when I can trace it back through the accounts and out to the bank statements without a translator. Owner add-backs, related-party rent, a personal car on the company books: these are normal in small and medium-sized enterprises, and none of them are disqualifying on their own. What matters is that every one of them can be explained and reconciled. If the earnings cannot survive due diligence, they were never earnings.

Add-back

An expense added back to profit because it belongs to the owner, not the business: a personal car, above-market rent to a related party, a salary the next owner will not pay. Normal in small companies, and fine as long as every one can be explained and traced.

Customer concentration

How much of the revenue leans on a single client. The higher it climbs, the more you are buying one relationship instead of a business.

I score this line hard because it is the one most often dressed up. A seller who has decided to exit has had a year to make the numbers look their best, and the honest way to read them is to assume the good year is the ceiling, not the floor. If I cannot get from the tax return to the bank feed and back, I treat the gap as risk, not detail. This is where having run a business myself earns its keep: I know which line items get quietly deferred to make a month look better than it was.

Durability, a reason to exist in twenty years

Profit tells me the business works now. Durability tells me it will still work after I own it. My rough test is age plus demand. I lean toward companies that are ten years or older, because a business that has already survived a recession, a bad hire, and a lost key account has proven something a five-year-old cannot. Age is not magic, but it is evidence: the thing has been shot at and is still standing.

Then I ask what the demand is made of. Recurring revenue, a maintenance contract, a part that wears out and gets reordered, a service a customer cannot skip: that is demand with a floor under it. The strongest question I know is the simple one. Is there a clear reason this business still exists in twenty years? If the answer depends on a trend holding or a platform staying kind, the durability is borrowed, not owned. I would rather own a plain demand I can see the bottom of than a fashionable one I cannot.

I am not buying last year's profit. I am buying the next twenty years of it.

Owner dependence is a discount, not a dealbreaker

Here is the part most buyers get backwards. A business that runs entirely through its founder, where he holds every relationship, quotes every job, and carries the whole operation in his head, is riskier. It is also cheaper. Owner dependence pulls the price down, and that discount is exactly where a buyer like me makes his return.

The margin sits in the gap between how the business runs today and how it could run with an operating layer installed underneath it: documented processes, a second person who can quote, systems that hold the knowledge instead of one tired human. I am not paying a premium for a business that is already clean. I am paying a discount for one that works despite its dependence, then closing that gap myself.

The risk is real and I price it honestly. A founder who is the entire product is a different animal from a founder who is merely the current bottleneck, and the whole job in diligence is telling those two apart. If the dependence is a set of habits and relationships I can document and transfer, it is a discount. If it is a rare licence, a single irreplaceable skill, or a personality the customers are actually loyal to, it is not a discount, it is the business, and it walks out the door with the seller. I go deep on this one, because it is where I both make my money and lose it. I wrote the long version in the owner-dependence piece.

Boring wins

The businesses that clear this filter are almost never exciting. Service companies. Trades. Niche industrial suppliers. Quiet business-to-business operations that have sold the same unglamorous thing to the same customers for two decades and will keep doing it long after the flashy names have pivoted three times.

Boring is a feature. A business nobody writes about is a business nobody is racing to disrupt or bid up. The demand is steady because the need is dull and permanent. Somebody has to fix the thing, supply the part, run the route, keep the system alive. The trade-off is real, and I take the same side every time: I would rather own a plain company with a floor under its revenue than a bright one whose whole value lives in a story that has to keep being retold.

The objection I hear is that boring means slow, and slow means leaving money on the table. Maybe. But I am not trying to win the year, I am trying to own something for twenty of them. A dull business with steady demand compounds quietly and does not need me to defend a narrative every quarter. That is exactly the kind of thing I want to buy and hold.

What I pass on fast

The filter is as much about the fast no as the slow yes. Three things end the conversation quickly, and each one has the same shape: a tell you can spot early, a reason it kills the deal, and a single narrow exception that proves the rule.

A turnaround dressed as a bargain

The tell is a price that looks too good next to the story. The business is cheap, and when you ask why, the answer is some version of "it just needs the right owner." That is not a discount, it is a repair bill the market has already priced in.

It kills the deal because I buy things that already work, not projects that need rescuing before they earn a cent. A broken business asks me to be right about the fix and to have the time, the cash, and the patience to execute it while it bleeds. That is a different sport from the one I am playing, and it fails the first line of the buy-box: earnings that survive a cycle.

The one exception is a business that is sound but sits inside a distracted owner, not a broken one: profitable, wanted, simply neglected because the founder checked out years ago. That is not a turnaround, that is owner dependence wearing a tired coat, and it belongs in the discount pile, not the trap pile.

Customer concentration

The tell is one logo doing too much of the work. You ask for revenue by customer and the top line is a third, a half, more. Sometimes the seller volunteers it as a strength, a marquee client, a long relationship. Read it the other way.

It kills the deal because if one client is a large share of the revenue, I am not buying a business, I am buying a single relationship I did not build and cannot control. The day that customer leaves, the earnings leave with them, and they tend to leave precisely when ownership changes hands and the person they trusted is gone. That is why the buy-box caps any one customer at a fifth.

The one exception is concentration that is structural and sticky rather than personal: a supplier wired into a customer's production line, switching costs measured in months of disruption, a contract with real teeth and time left to run. Even then I treat it as a reason to pay less and to structure the deal around the risk, never as a detail to wave through.

Books nobody can explain

The tell is a story that shifts under pressure. You ask a clean question and get a cloudy answer; you ask it again next week and the shape of the number has changed. The owner waves off the reconciliation, or the management accounts and the tax return tell two different tales.

It kills the deal because what I cannot verify, I will not buy. It is not always fraud. Sometimes it is just chaos, a business run out of a founder's memory with the paperwork treated as an afterthought. But chaos I cannot diligence is indistinguishable from a problem being hidden, and I am not paid to guess which one I am looking at.

The one exception is messiness that resolves under work: the numbers are disorganised but, given time and the source documents, they reconcile cleanly and the story stops moving. Small businesses are rarely tidy. The line I hold is not tidy versus untidy, it is verifiable versus not.

I buy what already works, then install the layer that makes it work without me. The filter exists to make sure I am adding upside, not life support.

This is the buy

Everything above serves one idea. I buy what already works, then install the operating layer on top of it. The filter is designed to find a business that does not need me to survive, so that what I add is upside rather than life support. The operating layer is the actual work: systems, process, a second person who can quote, knowledge that lives in the company instead of one head.

Picture the kind of company that clears every line: a two-truck industrial parts supplier, fifteen years old, boring as a Tuesday, run almost entirely by a founder who wants to retire and knows every customer by name. On paper it is fragile because it leans on him, and that fragility is the price coming down. Say its biggest customer sits under a fifth of revenue, the books tie out to the bank, and the demand is a part that wears out and gets reordered forever. That business is not a gamble. It is a discount waiting for someone to install the layer the founder never had time to build. It is hypothetical, but it is the exact silhouette the buy-box is cut to find.

This filter is the discipline that keeps the pipeline honest. I am writing this as documentation, as I go. A business worth buying is one that already works, priced for the gap between how it runs and how it could. Find that, install the layer, and hold.